

*** Replication .do file for Measuring Emotional Response: Comparing Alternative Approaches to Measurement 
*** Authors: Marcus, George; Williams College, Political Science
	*Neuman, W.; New York University, Communications
	*MacKuen, Michael; University of North Carolina Chapel Hill, Political Science
*** Stata 12 or above


log using "Marcus_et_all.txt", replace
use "AdultIVRecoded5.dta"

*Page 15: Chronbach�s Alpha - three item interest in new information
alpha google searchweb challengeme if Method2 <3

*Page 16: correlations
correlate nocompromise rightone if Method < 3
correlate resolveeveryone everyoneconcerns if Method < 3
correlate nocompromise rightone resolveeveryone everyoneconcerns if Method < 3

*Page 17 [table 1] : means and missing data for ten affect items x methodby Method2, sort : summarize scared enthusiastic hateful worried hopeful angry proud bitter resentful afraid if Method2 < 3
by Method2, sort : summarize scared enthusiastic hateful worried hopeful angry proud bitter resentful afraid if Method2 < 3

*Page 17: time to complete and t-test of two methods
by Method2, sort : summarize sliders_timing if Method2 < 3
by Method2, sort : summarize qradio_timing if Method2 < 3

ttesti 134 103 192.27 154 85 143.34

*Page 18: Box plots of ten affect items by method
graph box hopefulRB proudRB enthusiasticRB angryRB hatefulRB bitterRB resentfulRB afraidRB scaredRB worriedRB if Method2==0
graph box hopefulSL proudSL enthusiasticSL angrySL hatefulSL bitterSL resentfulSL afraidSL scaredSL worriedSL if Method2==1

*Page 19: factor analysis - by method - of ten affect items (add scree plot)
factor scared worried afraid enthusiastic hopeful proud hateful angry bitter resentful if Method==1 [aweight = weight], factors(3)
rotate, promax oblique

factor scared worried afraid enthusiastic hopeful proud hateful angry bitter resentful if Method==2 [aweight = weight], factors(3)
rotate, promax oblique

*Page 20: SEM models (first as distinct models then testing whether there are significant differences between the two - i.e., a model effect).
*First each method Separate (as in Figure 4)

sem (Enthusiasm -> hopeful, ) (Enthusiasm -> proud, ) (Enthusiasm -> enthusiastic, ) (Aversion -> angry, ) (Aversion -> hateful, ) (Aversion -> bitter, ) (Aversion -> resentful, ) (Anxiety -> afraid, ) (Anxiety -> scared, ) (Anxiety -> worried, ) if Method==1, covstruct(_lexogenous, diagonal) standardized latent(Enthusiasm Aversion Anxiety ) cov( Enthusiasm*Aversion Enthusiasm*Anxiety Aversion*Anxiety) nocapslatent
estat gof, stats(all)


sem (Enthusiasm -> hopeful, ) (Enthusiasm -> proud, ) (Enthusiasm -> enthusiastic, ) (Aversion -> angry, ) (Aversion -> hateful, ) (Aversion -> bitter, ) (Aversion -> resentful, ) (Anxiety -> afraid, ) (Anxiety -> scared, ) (Anxiety -> worried, ) if Method==2, covstruct(_lexogenous, diagonal) standardized latent(Enthusiasm Aversion Anxiety ) cov( Enthusiasm*Aversion Enthusiasm*Anxiety Aversion*Anxiety) nocapslatent
estat gof, stats(all)

* Second group analysis

sem (ANXIETY -> scared) (ANXIETY -> worried) (ANXIETY -> afraid) (AVERSION -> angry) (AVERSION -> hateful) (AVERSION -> bitter) (AVERSION -> resentful) (ENTHUSIASM -> enthusiastic) (ENTHUSIASM -> hopeful) (ENTHUSIASM -> proud) if Method < 3 [pweight = weight], covstruct(_lexogenous, diagonal) group(Method) latent(ANXIETY AVERSION ENTHUSIASM ) cov( ANXIETY*AVERSION ANXIETY*ENTHUSIASM AVERSION*ENTHUSIASM) nocapslatent
sem, standardized
estat ggof


* Page 21-22 [table 2] - Chronbach�s alpha and Feldt test.
*Anxiety
by Method, sort : alpha scared worried afraid

*Enthusiasm
by Method, sort : alpha enthusiastic hopeful proud

*Aversion
by Method, sort : alpha hateful angry bitter resentful



*Page 22: Anova�s for Anxiety, Enthusiasm, and Aversion (w interaction term for MethodxStory)
*For Anxiety
anova ANXIETY Story Method Method#Story

* For Enthusiasm
anova ENTHUSIASM Story Method Method#Story

*For Aversion
anova AVERSION Story Method Method#Story


*Page 22 [figures 5-7] plots of means by story x method

*for Anxiety (Figure 5)
mean ANXIETY, over(Method2 Story)


*for Enthusiasm (Figure 6)
mean  ENTHUSIASM, over(Method2 Story)

*for Aversion (Figure 7)
mean AVERSION, over(Method2 Story)

*Page 24: Regression - novel information - Table 3
by Method2, sort : regress ATTENTION3 ANXIETY ENTHUSIASM AVERSION PartyID Ideology NeedforCognition Extraversion Agreeableness Conscientiousness Neuroticism OpennessExperience [aweight = weight]

*Page 24: Regression - SCR - Table 4
by Method2, sort : regress StandPat2 ANXIETY ENTHUSIASM AVERSION PartyID Ideology NeedforCognition Extraversion Agreeableness Conscientiousness Neuroticism OpennessExperience [aweight = weight]

*Page 24: Regression - SOM - Table 5
by Method2, sort : regress Cooperate2 ANXIETY ENTHUSIASM AVERSION PartyID Ideology NeedforCognition Extraversion Agreeableness Conscientiousness Neuroticism OpennessExperience [aweight = weight]

*Page 24, footnote 15 - Interest in Novel Information
by Method2, sort : regress ATTENTION3 ANXIETY ENTHUSIASM AVERSION [aweight = weight]


log close


